This project implements a self‑learning face recognition system using OpenCV’s LBPH algorithm. It can detect new users, register them automatically, recognise returning users and optionally log attendance.

• dataset/ — folder where captured face images are stored. Each user’s images are saved as User.<id>.<count>.jpg.
• labels.json — JSON file mapping user IDs to names.
• trainer.yml — trained LBPH model. Generated automatically when the system retrains.
• smart_recognize.py — main Python script that performs real‑time detection, registration and recognition.
• attendance.csv — optional log of recognised users with timestamps.

• Python 3.8+
• Packages: opencv-contrib-python (not opencv-python), numpy
• A webcam

Install dependencies with:

pip install opencv-contrib-python numpy

Run the main script from this folder:

python smart_recognize.py

The program opens a webcam window. When a face appears:

• If it recognises the face, it displays a welcome message and logs attendance.
• If it does not recognise the face, it prompts you to press n to register a new user. Type the name when prompted. The system automatically captures 30 images of the user’s face for training and retrains the model.

Press q to quit at any time.

Refer to the accompanying guide for more details.